iN NATURAL CLASSIFICATION. 585 
The alterations in figure, which the pollen undergoes by 
means of re-agents, have been long since noticed. 
As regards the structure of the pollen, we possess no cha- 
racters of high value. Fritzche classifies this organ, accord- 
ing as the grains are possessed of one, two, three, or four 
membranous envelopes. But, in the family of the Fluviales, 
Zostera marina and Najas major have only a single integu- 
ment (the Intine), while Ruppia maritima, likewise, has an 
Exine ; and the existence of many integuments separates the 
Onagrariacee from all their allies. I have shown, in the 
Journal of Botany for June, 1841, that the exine frequently 
consists of two parts, a proper membrane and certain appen- 
dages upon its surface. The pressure of these latter was 
regarded with undue respect by Guillemin, who considered 
their existence of sufficient importance to divide not only 
families, but classes from each other. But I find that Cen- 
taurea has a smooth pollen, while this organ in the remain- 
ing Composite is hispid ; and those of the Ranunculacee that 
have spherical pollen, have the exine coated with append. 
ages, while in those with elliptical pollen, itis smooth. I 
may remark in this place, nevertheless, that in a practical 
point of view, the translucency or opacity of the exine, is a 
character of high value. 
Another character that is worthy of remark, is the com- 
pound nature of some pollen. 1 do not here allude to what 
are called pollen masses in the Orchidee. These consist of 
the * pollen grains," properly so called, united together by a 
cellular connection. It is in the Listera ovata, or some of 
those species of Orchidee, that are said to have * powdery 
pollen,” that the true nature of this organ can be best 
examined, where it will be found to consist of three, four, or 
five vesicles united together; these are rendered much more 
distinct by nitric acid, in which the exine peels off. If the 
pollen of Orchis latifolia, or maculata, be examined when the 
flower is very small (the spire being not more than one- 
fourth of the length from the summit of the ovary to the 
Apex of the unexpanded sepals), it will be found to consist of 
2X 
